You are on page 1of 32

UNDERSTANDING AND PREVENTING CORROSION IN

ALKANOLAMINE GAS TREATING PLANTS

Michael S. DuPart

Thomas R. Bacon

David J. Edwards

Dow Chemical USA

GAS /==== Technology Group


====/ SPEC

Freeport, Texas

41st Annual Laurance Reid


Gas Conditioning Conference

March 4 - 6, 1991

Norman, Oklahoma
UNDERSTANDING AND PREVENTING CORROSION IN
ALKANOLAMINE GAS TREATING PLANTS

Michael S. DuPart
Thomas R. Bacon
David J. Edward

Dow Chemical USA - GAS /==== Technology Group


====/ SPEC

INTRODUCTION

Corrosion in alkanolamine gas treating plants is virtually


impossible to eliminate. It can, however, be controlled and
even minimized. Corrosion can result in unscheduled down time,
loss of production, reduction in equipment life and even
injury or death. It is critical that design engineers and
plant operators understand the impact that corrosion can have
on an amine plant. In this paper we will discuss corrosion
mechanisms, different kinds of corrosion that can occur in
amine units, contributing factors to corrosion and
preventative measures. Emphasis will be placed on preventative
measures such as design considerations, operating parameters
and solvent choice. In addition, corrosion case histories in
amine treating will be reviewed.

BACKGROUND

The use of a regenerative process employing alkanolamines for


C02 and H2S removal dates back to the early 1930's. It was
recognized early on that removing H2S and/or C02 with
alkanolamine based gas conditioning solvents posed a problem
from a corrosion standpoint. The amine itself was not the
culprit, but the acid gas that the amine had absorbed. Several
investigators have shown that carbon steel when exposed to
alkanolamines under an inert gas pad exhibited lower corrosion
rates than when compared to water solutions under similar
conditions.

Extensive corrosion data has been generated that shows the


effect of acid gas on corrosion rates with alkanolamines. As
illustrated in Figure 1 for MEA, not only does the acid gas
have an effect, but the type of acid gas and the ratio of H2S
to C02 impacts the corrosion rate as well. The data indicates
that MEA with H2S or C02 alone is more corrosive than a mixture
of the two. In Figure 2 the effect of acid gas loading is
shown and demonstrates that as the loading is increased,
corrosion rates increase accordingly.2 As temperature is
increased, corrosion rates also increase and this trend is
shown in Figure 3 as well as in the previous figures. Similar
data has been generated for DEA on the effects of acid gas and
temperature as shown in Figure 4. It should be pointed out
that there appears to be no satisfactory correlation available
for predicting corrosion behavior with mixtures of H2S and C02
in either MEA or DEA service. However, plant data indicates
that in predominantly C02 serviced small quantities of H2S have
a strong passivating influence.

The DEA corrosion data in Figure 4 closely parallels the MEA


data. DEA is generally considered to be a less corrosive amine
than MEA, but only in relative terms as the data indicates.
MDEA / C02 only corrosion testing of carbon steel indicates
that there may be an advantage to the use of MDEA from a
corrosion standpoint. Figure 5 shows comparison corrosion
results of MEA, DEA, and MDEA where the MDEA has significantly
lower corrosion rates than either MEA or DEA. Plant operating
data seems to confirm the MDEA laboratory corrosion test
results. Figure 5 also reveals the tendency of higher
concentration amines to be more corrosive. While MDEA is not
immune to corrosion, the success of MDEA can be attributed to
its inherently lower corrosivity and service in mixed acid gas
applications which has been shown to reduce corrosion
tendencies. DEA has also exhibited good corrosion service in
mixed acid gas situations.

Concerns about corrosion in alkanolamine gas treating units


are generally focused on the rich side of the cross exchanger,
rich amine piping after the cross exchanger, in the still and
in the reboiler due to the overwhelming evidence that free
acid gas and higher temperatures are the main driving forces
for corrosion.

TYPES OF CORROSION

While a large number of corrosion failures in alkanolamine


service can be attributed to free acid gas and high
temperature, other contributing factors have accounted for
numerous corrosion experiences. A discussion of the different
types of corrosion will reveal other means by which corrosion
can occur.

General Corrosion

General corrosion, the most widely recognized form, is


characterized by uniform material deterioration over the
entire exposed area. Although this corrosion type destroys a
great deal of material annually, equipment life can be
accurately monitored and predicted by relatively simple
corrosion tests. These tests include in-line corrosion probes
and coupons or ultrasonic thickness testing. General corrosion
rates have been established for amine treating plants based on
more than fifty years of operating experience with different
types of metallurgy with corrosion allowances set accordingly.

Unfortunately, most corrosion failures in amine service are


not related to general type corrosion, but localized attack.
Localized corrosion usually occurs in specific areas and goes
undetected until unexpected or premature failure occurs.
Several forms of localized corrosion are discussed below.
Galvanic Corrosion

When two dissimilar metals are coupled together through a


conducting electrolyte, an electrical potential is set up
between the two metals which causes the less resistant metal
to corrode preferentially over the other. Galvanic corrosion
can occur, for example, when stainless steel trays are held
together with carbon steel bolts. The small carbon steel
bolts, acting as electrical anodes, would be preferentially
corroded over the more resistant stainless steel tray because
of the difference in electrical potential between the two
metals. Another example of galvanic corrosion is pump shafts
and valve stems made of steel or more corrosion resistant
material failing because of contact with more noble graphite
packing. Area effects are also important when considering
galvanic corrosion.6 In the example used above if carbon steel
trays were held together with stainless bolts, corrosion would
be minimized due to the very large anode (tray) and the small
cathode (bolt).

Crevice Corrosion

Crevice corrosion results when a concentration cell is formed


in the crevice between two metal surfaces or a metal and a
non-metallic material. Crevice corrosion can be caused by
oxygen depletion, changes in acidity or by depletion of an
inhibitor in the crevice.6 The presence of chlorides ions can
contribute greatly to crevice corrosion. Crevice corrosion is
frequently found in amine treating plants at the points where
tubes are attached to a tube sheet in a heat exchanger and is
more common in the reboiler.8 One option to prevent this type
of crevice corrosion is to weld the tubes to the tube sheet
rather than roll them in.6 Crevice corrosion can also occur
under deposits in stagnant areas and filtration is recommended
to prevent solid deposition.
Pitting Corrosion

Pitting corrosion is a highly localized and intense type of


corrosion, resulting in rapid and destructive penetration at
relatively few spots. Equipment failures due to pitting
corrosion are often quite sudden and unexpected. The presence
of halide ions, particularly chlorides, as well as stagnant
areas usually promote pitting type corrosion. For plants
employing corrosion inhibitors, inhibitor concentration is
critical to controlling this type of corrosion. If the pitting
is not completely stopped by the inhibitor, the intensity of
the pitting may be increased.6 It is interesting to note that
carbon steel is more resistant to pitting corrosion than
stainless steel.

Intergranular Corrosion

Corrosion that occurs selectively at metallic grain boundaries


is known as intergranular corrosion. This type of corrosion is
most frequently found in stainless steels. Precipitation of
chromium carbide in grain boundaries by improper heat
treatment or welding techniques promotes corrosion via the
intergranular route. An excellent example of this type of
corrosion is outlined in Case Number 5 in the "Case Histories"
section at the end of the paper. Heat treatment outside the
sensitizing range and development of low carbon stainless
steels such as 304L and 316L have helped to reduce
intergranular type corrosion. Two other forms of intergranular
corrosion, weld decay and knife line attack, are the result of
improper welding and lack of post weld heat treatment.
Electric arc welding, as opposed to gas welding, is
recommended to prevent weld decay and post weld heat treatment
for knife line attack of stabilized stainless steels.6

Selective Leaching

Selective leaching is the removal of one element from a


metallic alloy by a corrosion like process. A classic example
of this type of corrosion in amine systems is the selective
leaching of copper out of Monel or 70/30 Cu/Ni when oxygen is
present in the system.8 Generally, selective leaching is not a
concern in amine treating plants due to the infrequent use of
copper bearing alloys and the absence of oxygen from most
systems.

Erosion Corrosion

The acceleration or increase in rate of deterioration or


attack on a metal because of relative movement between a
corrosive fluid and the metal surface is defined as erosion
corrosion.6 This type of corrosion is aggravated by the pres-
ence of solids particles or gas bubbles suspended in the
liquid, both of which are commonly found in amine solutions.
Erosion corrosion is prominent in plumbing bends, elbows, tees
and valves as well in pumps, blowers and impellers. Velocity,
turbulence and the stability of protective films also effect
the degree of erosion corrosion. Impingement attack (corrosion
caused by turbulent flow} and cavitation damage (caused by the
formation and collapse of vapor bubbles in a liquid near the
metal surface) are two examples of erosion corrosion.6

Stress Corrosion Cracking

Stress corrosion cracking results from the simultaneous


presence of tensile stress and a corrosive media. Stress can
be either residual internal stress in the metal, or an
externally applied stress.6 Chloride induced stress corrosion
cracking is one of the most widely recognized form of this
corrosion type. Other factors that affect the extent of stress
corrosion cracking are the operating temperature, chemical
composition of the amine solution, metal composition and
structure.

Stress corrosion cracking in amine units has received


considerable attention in recent years after the catastrophic
rupture of an MEA contactor at a U.S. refinery that resulted
in seventeen fatalities and extensive property damage. Even
though this failure was later attributed to hydrogen induced
cracking, a joint API/NACE stress corrosion cracking survey of
amine plants was conducted after this incident. The survey
results are quite extensive, but seem to indicate that stress
corrosion cracking is most prevalent in DEA units and to a
much lesser degree in DEA and MDEA units.10

Hydrogen Damage

Hydrogen damage is a general term which refers to mechanical


damage of a metal caused by the presence of, or interaction
with hydrogen.6 In amine service hydrogen damage is usually
limited to either blistering, embrittlement or attack. Amine
solutions containing hydrogen sulfide or cyanide can promote
either hydrogen embrittlement or sulfide stress cracking.6,10
Hydrogen blistering can occur in corrosive amine solutions
that contain hydrogen evolution poisons such as sulfides,
arsenic compounds, cyanides and phosphorous containing ions.6
Hydrogen associated failures are often mistaken for stress
corrosion cracking and this may account for the high rate of
stress corrosion cracking incidents reported in MEA service.
MECHANICAL DESIGN CONSIDERATIONS

There are several mechanical design criteria, when implemented


into an amine plant design, can minimize corrosion. Some of
the criteria are based on common sense, but others have been
developed based on plant operating experience.

Metallurgy

Ideally, from an upfront capital cost perspective, one would


like to fabricate an amine plant solely from carbon steel.
Plants have operated successfully without corrosion problems
made entirely out of carbon steel, but the process concessions
required to do so result in high energy consumption.
Substituting stainless steel in key areas of the plant (rich
amine piping, cross exchanger tubes and internals of the
reboiler and stripper) will provide adequate corrosion
protection and allow the plant to be operated with optimum
energy input. Table 1 and Figure 6 show corrosion rates for
different amines with various alloys under both laboratory and
plant conditions.11,12,13 Carbon steel, as indicated by the data,
exhibited corrosion rates, depending on the test conditions,
from <1 mpy to >100 mpy. Stainless steel, in particular 316,
performed adequately under all test conditions. 410SS has
shown to be anodic to carbon steel in some applications and
caution should be exercised when specifying this material.
While some investigators have shown acceptable corrosion rates
with Monel and other copper bearing alloys , its use should be
discouraged due to unfavorable plant experience and evidence
that suggests corrosion of these alloys is enhanced by
oxidizers and amine degradation products. Table 2 lists
general metallurgical guidelines for amine plant construction
that have proved to be adequate. The use of dissimilar metals
should also be avoided to minimize bimetallic or galvanic
corrosion.14

Flow in Piping and Vessels

Minimizing acid gas flashing, that is keeping the C02 and/or


H2S in solution, is a major factor in controlling corrosion in
amine units. Several investigators have recommended 3
feet/second as the maximum solution velocity for flow in
carbon steel pipe.14,15,16 Solution velocities of 5 - 8
feet/second may be employed with stainless steel piping due to
its higher inherent resistance to corrosion. Other design
practices that will reduce turbulence and localized pressure
drop are use of long radius elbows and seamless pipe where the
solution changes direction, eliminating flow restrictions such
as flange gaskets protruding into the line, tubes that are not
flush with tube sheet, not using threaded connections
or socket weld fittings and using long tapered sections
(optimum cone angle of 7 0) when changing cross sectional area
of flow.16,17 Pump cavitation can be prevented by minimizing
suction pressure drop (reduced velocity, straight runs of
pipe, etc.) and providing sufficient NPSHA to prevent
formation of small gas bubbles.14,17

Control Valves

The location of control valves is important so that acid gas


flashing and corrosion is minimized. The valves controlled by
the level control in the bottom of the stripper or reboiler
should be downstream from the lean amine cooler.14 The
equivalent valve operated by the level control on the absorber
should be downstream from the lean/rich cross exchanger or if
a flesh tank is employed, between the absorber and exchanger.
14,16
Close attention to the design and of the control valve and
adjacent downstream piping is also recommended to reduce high
velocity two-phase flow associated with throttling.16

Filtration

Adequate mechanical and carbon filtration can significantly


reduce corrosion in amine units. Not only can filtration
reduce particles that promote erosion corrosion and abrasion
of passive films, it can effectively remove hydrocarbons and
to a lesser extent, amine degradation products.18 Filter
arrangements should remove all particles greater than 5
microns.14 In particularly dirty service, a spare parallel
mechanical filter should be incorporated into the design.
Caution should be exercised with paper and sock type filters
as they can collapse and recontaminate the solution.14
Mechanical filters should be placed both upstream and down
stream of the carbon bed. The upstream filter acts as a
"guard" filter and will extend carbon life. The downstream
filter prevents carbon fines from entering the system. In sour
service a full rich stream filter is recommended and in C02
only applications a slip stream filter on the lean stream is
adequate. Some plants have incorporated both rich and lean
filtering with particular success.

Stress Relieving

Stress relieving of amine plant hardware has been recommended


for many years. 14,19 Although cracking has occurred in vessels
and piping that were stress relieved, the NACE/API survey
showed that the vast majority of cracking occurred in
non-stress relieved equipment. Stress relieving should be
applied for all types of amine service and is particularly
important in MEA units due to their poor cracking record and
strong evidence that suggests they are specially prone to
cracking.20,21

Miscellaneous

There are a few other design considerations that can minimize


corrosion in amine units. Inlet knockouts are useful in
preventing brine from entering the amine solution. It has
already been pointed out how chlorides can promote pitting,
stress corrosion cracking and crevice corrosion. Inert or
sweet gas blanketing of the surge drum and storage tank will
minimize oxidation of the amine and prevent degradation
products from forming.14 DEA is particularly susceptible to
oxygen degradation and extra attention should be given to
excluding oxygen in DEA plants.22 Figure 7 illustrates this
point. Regeneration pressure should be kept as low as
possible with the bulk amine temperature not to exceed 255°F
and the reboiler heat source not to exceed 300°F. Operating
within these guidelines will minimize thermal degradation of
the amine. The reboiler can also have tubes removed in an "X"
or "V" pattern to minimize vapor binding and localized
overheating, a square pitch tube pattern to facilitate
cleaning and multiple vapor exit lines to prevent stagnant
areas of acid gas.14,16

AMINE & PROCESS PARAMETERS

Careful selection of amine and process parameters can have a


positive impact on reducing corrosion in amine gas treating
plants. Since "free" acid gas is the primary culprit
contributing to corrosion, a brief discussion regarding the
actual mechanism is appropriate.

The best explanation for C0 2 corrosion in amine units states


that reduction of undissociated carbonic acid in areas of
turbulence where a passive ferrous carbonate scale is unable
to be laid down is chiefly responsible for corrosion.23 The
carbonic acid can undergo either one of the following
reactions:

[1] H2CO3 + e ==> HCO3 + H (corrosion cathode)

[2] H2CO3 ==> CO2 + H2O (CO2 gas evolution)

The basic corrosion reaction for H2S is a little more straight


forward:

[3] Fe + H2S ==> FeS + H2

The corrosion reactions can get quite complicated in mixed


acid gas service and the variation in corrosion experience is
probably related to the difference in iron sulfide films which
can be formed, some are tightly adhering while others are
categorized as soft or jelly like.24

Choice of Amine

MEA, DEA and MDEA are the three most popular amines utilized
for acid gas treating (see Figure 8). Historically, MEA has
the worst reputation for corrosion problems. However, several
investigators have shown essentially no difference in
corrosion rates of the different amines in the absence of acid
gas.1,3,11,25,28 It is also well documented that MEA and DEA form
degradation products when reacted with CO2 and MDEA does
not.26,27,28 One investigator claimed that MEA/C02 degradation
products can enhance corrosion, but are not alone responsible
for corrosion.2 Other investigators have shown that DEA/C02
degradation products in the absence of acid gas do not enhance
corrosion. If MEA, DEA and MDEA have similar corrosion rates
with pure solutions and in the presence of basic degradation
products in the absence of acid gas, particularly C02, how does
one explain plant and laboratory corrosion results that show
MEA is more corrosive than DEA and DEA more corrosive than
MDEA? Teevens21 may have the best answer to this question. He
proposes that since MEA is a stronger Lewis base than DEA and
MDEA the weakest, reactivity plays a strong role:

"The formation of basic degradation products is directly


attributable to the reaction of C02 with primary amines and
secondary amines which react to form intermediate amides,
and, subsequently, their amine-carbamates which in turn,
undergo internal dehydration to yield various basic
degradation products. Since MEA is a much stronger Lewis
base, then it is likely to form more carbamates than DEA.
The production of carbamates from tertiary amines is
impossible, since they fail to yield amides necessary in
carbamate ion formation. The presence of amine carbamates
as viable intermediary salts, may, somehow impact the
corrosivity of MEA and DEA. The inability of tertiary
amines to react with C02 to form amides and subsequently
amine carbamates may be the vital clue as to why solvents
such as MDEA are indeed less corrosive in all aspects."

Operating MEA, DEA and MDEA plants have all demonstrated that
corrosion can be minimized under proper operating conditions.
However, based on plant experience and laboratory corrosion
data, relative corrosivity of the amines would be ranked as
follows: MEA >> DEA >> MDEA with MDEA the least corrosive. It
would appear that MDEA is the most forgiving amine from a
corrosion standpoint.
Amine Strength & Acid Gas Loading

Previous data presented indicated that as amine strength and


acid gas loadings are increased, corrosivity is enhanced.
Again this trend is supported by plant experience. Higher
strength amines also cannot achieve as high mole/mole
equilibrium rich loadings as compared to lower strength amines
and this can increase the potential for acid gas flashing.
Listed below in Table 3 are recommended maximum ranges for
amine strength and acid gas loadings that have proven
historically to adequately address corrosion concerns.

Table 3
_________________________________________________________
_
Amine Wt% Rich Loading, M/M
________________________________________________________
__
MEA 15 - 20 0.30 - 0.35
________________________________________________________
__

DEA 25 - 30 0.35 - 0.40


________________________________________________________
__

MDEA 50 - 55 0.45 - 0.50


___________________________________________________________
_

Consideration should also be given to maximum lean loadings


and listed in Table 4 are recommended loading in the lean
circuit to minimize acid gas flashing. These loadings should
be easily achieved with a 1.0 - 2.0 M/M stripper reflux ratio.

Table 4
_________________________________________
Amine Total Lean Loading, M/M
__________________________________________
_
MEA 0.10 - 0.15
__________________________________________
_
DEA 0.05 - 0.07
__________________________________________
_
MDEA 0.004 - 0.010

Heat Stable Salts

Anions such as formate, oxalate, thiosulfate, thiocyanate and


chloride that tie up the amine and are not capable of being
regenerated in the stripping section of the plant are defined
as heat stable salts (HSS). While there is conflicting data on
the effect of HSS on corrosivity, the general consensus is
that HSS will enhance corrosion in alkanolamine gas treating
solutions. For this reason and because HSS also reduce the
acid gas carrying capacity of the amine they should be
minimized. Table 5 lists common sources for the anions and can
be used as a guide for minimizing HSS formation. As a general
rule of thumb, HSS should not exceed 10% of the total amine
concentration

Corrosion Inhibitors

Corrosion in amine treating units has been a problem for many


years and considerable research has gone into developing
inhibitors, particularly for high strength amine systems. The
most affective inhibitors developed, however, were based on
heavy metals such as arsenic and vanadium and their popularity
is waning due to the environmental concerns associated with
heavy metals. In addition, the heavy metal inhibitors while
providing excellent corrosion protection of wetted areas, did
not always adequately protect splash and vapor regions. "Off
the shelf" corrosion inhibitors based on film forming amines
have been employed in amine systems with limited success due
to their tendency to address general corrosion only without
adequately covering other forms of corrosion. Inhibitors are
also frequently hard to maintain and if not properly
controlled can lead to more severe corrosion. The best
approach to managing corrosion in amine units is thru proper
plant design and operation.

Make-up Water

Poor make-up water is often the source of chlorides and other


contaminants that will enhance corrosion. Using good quality
water cannot be overemphasized and listed in Table 6 are
minimum water quality standards.
Table 6
WATER QUALITY FOR AMINE PLANTS
________________________________________________

Total Dissolved Solids <100 ppm


__________________________________________
Total Hardness <3 grains/gal
__________________________________________
Chloride - <2 ppm
__________________________________________
Sodium <3 ppm
__________________________________________
Potassium <3 ppm
__________________________________________
Iron <lO ppm
__________________________________________
CASE HISTORIES

Case Number 1

Problem: Pitting and erosion corrosion in high loaded rich


formulated MDEA at pressure letdown points.

Situation: Treating was required for 300 psig gas with 45% C02
and 20 ppm H 2S to produce fuel gas. Process design called for
high rich loadings created by recirculating semi-lean and
utilized partial stripping via 10 psig flash of the rich to
semi-lean stream. Final stripping was accomplished with a
fired reboiler serviced stripper. The plant was a grass roots
design with all new equipment. The plant metallurgy was
predominantly carbon steel with 304SS intended for flashing
service, but found only in the cross exchanger and its
immediate downstream piping and the circulating-pumps.

Within weeks after plant start-up, leaks appeared in the


liquid level control valve on the line transferring rich
solvent from the absorber to the flash drum, a similar valve
on the line transferring semi-lean solvent from the flash drum
to the stripper, in the inlet piping of the semi-lean solvent
circulating pump which takes suction from the flash drum and
feeds the stripper and booster pump feeding the absorber, and
in the valve adjusting semi-lean solution flow to the
absorber.

Observations: An internal inspection found the corrosion was


localized and characterized as pitting and erosion. Corrosive
effects were specific to the carbon steel metallurgy and
flashing conditions. The absorber bottom showed no corrosion
and still had mill markings on the wall. Amine solution
analysis indicated very high C02 loadings (>0.8 moles C02 /
mole solvent, and >100% equilibrium) in the rich and semi-lean
solutions. Corrosivity tests indicated a somewhat greater
corrosivity (7 vs. 1 mpy) of the plant solution versus virgin
solution. The circulating pump suction strainer was fouled
with rust and scale and created additional pressure drop
between the flash drum and pump suction. Piping design
utilized socket weld fittings and velocities in the problem
areas were 10 - 40 fps with the exception of 3 fps circulating
pump suction.

Conclusions: The mechanics of corrosion were diagnosed as:

1. Flashing C02 causing pitting due to the corrosivity


of the wet C02 and cavitation erosion upon bubble
collapse.
2. Erosion caused by the combination of corrosion and
abrasion at high velocity impingement points in the
system.

Corrections:

1. Change piping in flashing service from carbon steel


to 304SS and size for two phase flow.

2. Limit carbon steel piping velocity in the rich and


semi-lean circuits to 5 fps.

3. Limit flow disruptions causing turbulence by using


butt weld, long radius fittings and improved piping
geometry.

Case Number 2

Problem: Erosion corrosion in bottom of ammonia plant amine


contactor.

Situation: Treating synthesis gas at 350 psig with 17% C02 to


remove C02 to a target of 100 ppm. Conversion from inhibited
MEA to formulated MDEA of less basisity required increasing
mass transfer of the C02 from the gas to the amine. The amine
liquid level in the contactor was raised from approximately 4'
below the inlet gas distribution header to about 6" above the
header.

Observations: An internal inspection of the contactor one year


after the conversion found erosion patterns on the contactor
wall corresponding to the impingement location of the inlet
gas from the farthest holes from the center of the ladder type
distributor. The erosion had consumed a major portion of the
corrosion allowance for the vessel.

Conclusions: The mechanics of the corrosion was diagnosed as


turbulent interaction between the inlet gas and liquid surface
which prevented the normal passivation layer from forming.

Corrections: The farthest holes in the inlet distributor were


welded closed. The eroded areas were cleaned and filled with a
metal impregnated epoxy material for protection from future
attack. Subsequent annual inspections have verified no further
corrosion.

Case Number 3

Problem: Severe general and galvanic corrosion in bottom of


ammonia plant amine contactor.
Situation: Treating synthesis gas at 348 psia with 18% C02 to
<100 ppm with 25% MEA inhibited against corrosion with a heavy
metal inhibitor. Contactor employed thirty six twopass trays.
Oxygen was injected into catalyst vessel upstream of contactor
for conversion of CO to C02 in the synthesis gas. Dramatic
rise of iron level in amine solution after failure of catalyst
allowed 2000 - 3000 ppm oxygen to enter into bottom of
contactor with feed gas for 3 - 5 days.

Observations: An internal inspection found "leopard spot"


corrosion: dark circular passive areas surrounded by active
general and galvanic corrosion regions around the entire ID of
-the vessel in the-vapor area between the liquid level in the
bottom of the tower and the first contact tray. The liquid
level in the tower was maintained 12" below the inlet gas
distributor. Corrosion in the active areas had progressed up
to the corrosion allowance of the carbon steel vessel.
Corrosion of the same type, but to a lesser degree was also
found in the vapor region between the bottom five trays and
was not present by tray 31 (five from the bottom). The bottom
three tray downcomers in the vapor region also exhibited
similar corrosion.

Conclusions: The corrosion mechanism was diagnosed as


penetration of the passive iron carbonate film in the vapor
region via the following reaction:

4FeCO3 + O2 + 10H2O ==> 4Fe(OH)3 + 4H2CO3

Once the passive film had been compromised in a few areas,


further corrosion was assisted by galvanic action of active
and passive regions in close proximity.

Corrections: The bottom five trays were removed and the inlet
gas distributor turned upside down with the exit holes
pointed down. The liquid level in the contactor was
maintained above the inlet gas distributor and the corroded
area in subsequent operation of the plant to prevent direct
contact with the inlet gas and allow the inhibited MEA to
provide corrosion protection. The plant continued to utilize
the catalyst bed and oxygen injection in upfront processing,
but exercised caution in its operation. Later inspections
showed the corrective actions to have been successful.

Case Number 4

Problem: Pitting and erosion corrosion of stripper internals,


cross exchanger tubes and booster pump in formulated MDEA
plant.

Situation: Treating natural gas at 800 psig with 1.5% C02 to


a treated gas specification of 100 ppm in a grass roots
facility. Stripper internals were carbon steel trays and
304 SS valves. Cross exchanger arrangement based on rich
amine flow consisted of two shells in series with the tube
metallurgy carbon steel in the first and 316 SS in the
second. Lean amine C02 loadings were consistently in the
range of 0.15 - 0.20 M/M with stripper overhead temperature
below 180 F. Plant was operated in this mode for about a
year and a half. Plant design also incorporated a 10% slip
stream carbon filter in the rich circuit, preceded by a
sock filter. Solution corrosivity test indicated a
corrosion rate of 35 mpy.

Observations: An internal corrosion inspection of the stripper


revealed severe pitting and erosion corrosion of the carbon
steel tray decks and enlargement of the valve openings where
90% the valves were gone and piled up on each tray. Light
pitting was also evident on the liquid feed tray of the
absorber. Pitting corrosion to the point of penetration that
initiated on the shell side was also discovered on the carbon
steel cross exchanger tube bundle. Severe erosion corrosion
was also present on the booster pump impeller and case. Large
quantities of carbon granules were also found in the bottom of
the stripper and surge tank and on the shell side of both
bundles in the cross exchanger. Very little carbon was found
at all in the carbon filter vessel.

Conclusions: The corrosion mechanism was diagnosed as:

1. Flashing C02 in the cross exchanger caused by


excessive pressure drop and high lean loadings due
to plugging by the carbon and insufficient stripping
steam
that resulted in pitting type
corrosion.

2. Erosion corrosion of the booster pump impeller and


case cause by carbon solids circulating around in
the system.

3. Galvanic and erosion corrosion of stripper trays


decks due to 304 SS valves being coupled with carbon
steel
decks and carbon again circulating around the
system.

Corrections:

1. Stripper internals were replaced with 316 SS trays


and valves.

2. Carbon steel cross exchanger bundle was replaced with


316 SS tubes and baffle plates.

3. Support screen was installed in carbon bed-and


personnel trained on proper installation of carbon.
4. A full flow mechanical filter was installed downstream
of the existing slip stream carbon and sock filters.

5. Plant personnel were trained in proper stripper


operations (maintain sufficient reflux ratio to achieve
a
0.015 - 0.020 M/M lean C02 loading).

6. Plant switched solvent suppliers and put a new load of


formulated MDEA in the plant.

Case Number 5

Problem: Intergranular corrosion in heat affected zone of 304


SS stripper shell welds.

Situation-: Stripper of ammonia plant CO2 removal section was


fabricated as carbon steel for the lower courses and 304 SS
for the upper courses of the shell. The internals were
inspected one year after conversion from inhibited MEA to a
formulated MDEA solvent.

Observations: Corrosion was found in the heat affected zone of


longitudinal and circumferential welds of the 304SS shell.

Conclusions: The mechanics of corrosion were diagnosed as


intergranular corrosion caused by sensitization of the
stainless steel as a result of the fabrication techniques or
choice of metallurgy for the vessel. The shell apparently was
not a low carbon grade of stainless. The weld metal chosen was
not known. Sensitization, or carbide precipitation results in
depletion of chromium and lowering of corrosion resistance in
areas adjacent to the grain boundary. The heat affected zone
was previously protected by the heavy metal inhibitor package
which was dissolved by the formulated solvent. In effect, the
previous inhibitor protection compensated for the reduced
corrosion protection of the stainless in the heat affected
zone.

Corrections: A correct weld procedure was specified to


maintain the corrosion resistance of the 304 SS and the
affected welds repaired. Subsequent annual inspections
indicated no further attack.

SUMMARY

Corrosion in alkanolamine gas treating plants can be


controlled and minimized with proper plant design and
operating parameters and monitoring the operation frequently
for unplanned process and amine excursions. The key is to
minimize acid gas flashing and to understand other
contributing factors
that can enhance corrosivity of amine solutions.
Hopefully we have presented information that the
design engineer and operating personnel of amine
plants can utilize to minimize corrosion in new and
existing units.
REFERENCES

1. McNab, A.J., Treseder, R.S.; "Materials Requirement


For a Gas Treating Process", Materials
Performance, Vol. 10, No. 1, 1971.

2. Polderman, L.D., Dillon, C.P., Steele, A.B.;


"Degradation of MEA in Natural Gas Treating
Service", Oil & Gas Journal, May 16, 1955.

3. The Dow Chemical Company, "Gas Conditioning Fact


Book", 1962.

4. Kohl, A.L., Riesenfeld, F.C.; "Gas Purification",


3rd. Edition, Gulf Publishing Company, 1979.

5. The Dow Chemical Company, GAS/SPEC Technology


Group, Internal Corrosion Test Results, 1985.

6. Fontana, M.G., Greene, N.D.; "Corrosion


Engineering", 2nd. Edition, McGraw-Hill Book
Company, 1978.

7. Pearce, R.L., DuPart, M.S.; "What is an Amine


Inhibitor System?", 64th Annual GPA Convention,
1985.

8. Pearce, R.L., DuPart, M.S.; "Corrosion in Gas


Conditioning Plants - An Overview", N.A.C.E.
Corrosion Paper # 39, 1987.

9. Sedriks, A.J.; "Corrosion of Stainless Steels",


John Wiley & Sons, 1979.

10. Richert, J.P., Bagdasarian, A.J., Shargay, C.A.;


"Extent of Stress Corrosion Cracking in Amine
Plants Revealed by Survey", Oil & Gas Journal,
June 5, 1989.

11. Lang, F.S., Mason, J.F.; "Corrosion in Amine Gas


Treating Solutions", Corrosion, Vol. 14, No. 2,
1958.

12. The Dow Chemical Company, Internal Corrosion


Test Results, 1980.

13. The Dow Chemical Company, GAS/SPEC Technology


Group, Internal Corrosion Test Results, 1990.
REFERENCES CONTINUED

14. Campbell, J.M., "Gas Conditioning and Processing",


4th. Edition, Campbell Petroleum Series, Inc., 1981.

15. Connors, J.S.; "Designing Amine Treaters To


Minimize Corrosion", Proceedings of the 1973
Gas Conditioning Conference.

16. Dingman, J.S., Allen, D.L., Moore, T.L.;


"Minimize Corrosion in MEA Sweetening Units",
Proceedings of the 1966 Gas Conditioning
Conference.

17. Sheilan, M., Smith, R.F.; "Hydraulic Flow Effect


On Amine Plant Corrosion", Oil & Gas Journal,
November 19, 1984.

18. Chakma, A., Meisen, A.; "Activated Carbon


Adsorption of Diethanolamine, Methyl
Diethanolamine and Their Degradation Products",
Carbon, Vol. 27, No. 4, 1989.

19. Maddox, R.N.; "Gas And Liquid Sweetening",


2nd. Edition, Campbell Petroleum Series,
Inc., 1977.

20. Schutt, H.U.; "New Aspects of Stress Corrosion


Cracking in Monoethanolamine Solutions", N.A.C.E.
Corrosion Paper # 159, 1988.

21. Teevens, P.J.; "Toward a Better Understanding of


the Cracking Behavior of Carbon Steel in
Alkanolamine Sour Gas Sweetening Units; Its
Detection, Monitoring and How to Avoid It",
N.A.C.E. Corrosion Paper # 198, 1990.

22. The Girdler Corporation, Report to the U.S. Navy, 1950.

23. Treseder, R.S.; "Some Comments on Proposed


Mechanisms for Carbon Dioxide Corrosion",
N.A.C.E. Corrosion Paper # 284, 1984.

24. McFaddin, D.E.; "H2S and C02 Corrosion of Carbon


Steel in Natural Gas Processing Plants", Oil &
Gas Journal, December 13, 1951.

25. Froning, H.R., Jones, J.H.; "Corrosion of Mild


Steel in Aqueous Monoethanolamine", Industrial
and Engineering Chemistry, Vol. 50, No. 12,
December, 1958.
REFERENCES CONTINUED

26. The Dow Chemical Company, Gas Conditioning


Laboratory; "Amine Degradation", August 27,
1965.

27. Kennard, M.L., Melsen, A.; "Mechanisms and


Kinetics of Diethanolamine Degradation",
Industrial & Engineering Chemistry Fundamentals,
Vol. 24, 1985.

28. Blanc, C., Grall, M., Demarais, G.; "The Part


Played by Degradation Products in the Corrosion
of Gas Conditioning Plants Using DEA and MDEA",
Proceedings of the 1982 Gas Conditioning
Conference.